Early Storable Liquid-Propulsion Efforts

It would appear that the Caltech group had not made great progress on liquid propulsion as of July 1940 when it was joined by Malina’s roommate, Martin Summerfield, who completed work on his Ph. D. during 1941 in the physics department at Caltech. After joining Malina’s group, he went to the Caltech library, consulted the litera­ture on combustion-chamber physics, and found a text with infor­mation on the speed of combustion. Using it, he calculated—much in the fashion of Thiel at Kummersdorf—that the combustion chamber being used by the GALCIT team was far too large, result­ing in heat transfer that degraded performance. So he constructed a smaller chamber of cylindrical shape that yielded a 20 percent increase in performance. Von Karman believed that roughly 25 to 30 percent of the heat in the combustion chamber would be lost, based on information about reciprocating engines. The eminent aerodynamicist had therefore concluded that it would be impossi­ble for rocket engines to be self-cooling, restricting both their light­ness and length of operation. Summerfield’s calculations showed 146 these assumptions about heat transfer to be far too high, indicating Chapter 4 that it was possible for a self-cooling engine to operate for a sus­tained period. Subsequent tests confirmed Summerfield’s calcula­tions, and Malina learned about the technique of regenerative cool­ing from James H. Wyld of Reaction Motors during one of his trips back East.2

For the moment, the group worked with uncooled engines burn­ing RFNA and gasoline. Successive engines of 200, 500, and 1,000 pounds of thrust with various numbers of injectors provided some successes but presented problems with throbbing or incomplete ini­tial ignition, which led to explosions. After four months of efforts to improve combustion and ignition, Malina paid a visit to the Na­val Engineering Experiment Station in Annapolis in February 1942. There he learned that chemical engineer Ray C. Stiff had discovered in the literature of chemistry that aniline ignited hypergolically with nitric acid. Malina telegrammed Summerfield to replace the gasoline with aniline. He did so, but it took three different injector designs to make the 1,000-pound engine work. The third involved eight sets of injectors each for the two propellants, with the stream of pro­pellants washing against the chamber walls. Summerfield recalled

that after 25 seconds of operation, the heavy JATO units glowed cherry red. But they worked on a Douglas A-20A bomber for 44 suc­cessive firings in April 1942, the first successful operation of a liq­uid JATO in the United States. This led to orders by the army air forces (AAF) with the newly formed Aerojet Engineering Corpora­tion, which Malina and von Karman had helped to found.3

Aerojet did considerable business with the AAF and navy for JATO units during the war and had become by 1950 the largest rocket – engine manufacturer in the world, as well as a leader in research and development of rocket technology. Until Aerojet’s acquisition by General Tire in 1944-45, the rocket firm and the GALCIT rocket project maintained close technical relations. Although GALCIT/JPL was involved essentially with JATO work from 1939 to 1944, in the summer of 1942 the project began designing pumps to deliver liquid propellants to a combustion chamber instead of feeding the propel­lants by gas under pressure. By the fall of that year, project engineers were working on using the propellants to cool the combustion cham­ber of a 200-pound-thrust engine.4

Meanwhile, as a result of Stiff’s discovery, Truax’s group at Annapolis began using 1,500-pound JATOs burning nitric acid and aniline on navy PBY aircraft in 1943. Both Truax and Stiff subse­quently got orders to work at Aerojet, where Stiff devoted his efforts to a droppable JATO using storable, hypergolic propellants. Aerojet produced about 100 of these units, and some came to be used by the U. S. Coast Guard. In these ways, Aerojet became familiar with use of storable propellants, and Stiff joined the firm after completing his obligatory service with the navy.5

The next important development involving engines with stor­able propellants was the WAC Corporal sounding rocket (the term WAC standing for Women’s Auxiliary Corps or Without Attitude Control, depending upon the source consulted). The Army Ord­nance Corps had requested that Malina’s project investigate the fea­sibility of developing a rocket carrying meteorological equipment that could reach a minimum altitude of 100,000 feet. The JPL team redesigned an Aerojet motor that used monoethylene as a fuel and nitric acid mixed with oleum as an oxidizer. The original motor was regeneratively cooled by the monoethylene. JPL adapted the motor to use RFNA containing 6.5 percent nitrogen dioxide as oxidizer and aniline containing 20 percent furfuryl alcohol as a fuel, thereby increasing the exhaust velocity from 5,600 to 6,200 feet per second but leaving the thrust at 1,500 pounds for 45 seconds. According to one source, the specific impulse was 200 lbf-sec/lbm (slightly lower than the V-2).6

Besides exceeding the requirements of the army, the small, liquid-propellant rocket also functioned as a smaller test version of the Corporal E research vehicle, providing valuable experience in the development of that larger unit. During the testing, the pro­gram decided to modify the WAC Corporal to attain higher alti­tudes. A substantial modification of the engine reduced its weight from 50 to 12 pounds. The WAC A initial version of the rocket had a comparatively thin, cylindrical inner shell of steel for the combus­tion chamber, with an outer shell that fit tightly around it but was equipped with a joint to permit expansion. Helical coils (ones that spiraled around the outside of the combustion chamber like a screw thread) provided regenerative cooling, with a shower-type injector in which eight fuel streams impinged on eight oxidizer streams. For the modified WAC B engine, designers reduced the combustion chamber in length from 73 to 61 inches and made minor modifica­tions to the injector. It had an inner shell spot-welded to the outer shell, still with helical cooling passages. The injector remained a showerhead with eight pairs of impinging jets.7

In a series of flight tests at White Sands Proving Ground, New Mexico, in December 1946, none of the WAC Corporal B vehicles rose more than 175,000 feet in altitude. Apparently the test team suspected cavitation (gas bubbles) in the injector system as the cause of the less-than-optimal performance, since team members 148 constructed three more B-model vehicles with orifice inserts that Chapter 4 were screwed in, rather than drilled as before, to achieve cavitation – free injection of the propellants into the combustion chamber. In three February-March 1947 tests, one WAC Corporal B reached an altitude of 240,000 feet. Overall, the WAC Corporal demonstrated that the propulsion system was sound and the nitric acid-aniline – furfuryl alcohol propellant combination was viable.8

The WAC Corporal led directly to the successful Aerobee sound­ing rocket built by Aerojet, which was used by the Applied Physics Laboratory of Johns Hopkins University for research in the upper atmosphere. Then, in the Bumper-WAC project, the WAC Corporal B flew as a second stage on V-2 missiles. The reported altitude of 244 miles and maximum speed of 7,553 feet per second (reached on February 24, 1949) were records. This highly successful launch demonstrated that a rocket’s velocity could be increased with a sec­ond stage and that ignition of a rocket engine could occur at high altitudes.9

In addition, the engine for the WAC Corporal contributed to the Corporal missile’s propulsion system. As first conceived, Corporal E was a research vehicle for the study of guidance, aerodynamic, and

propulsion problems of long-range rockets. In 1944, von Karman estimated that a rocket with a range of 30 to 40 miles would be necessary to serve as a prototype for a later missile. He thought such a vehicle would need an engine with 20,000 pounds of thrust and 60 seconds of burning time. Experience at JPL to that point had indicated that the only already-developed rocket type meeting von Karman’s specifications would be a liquid-propellant vehicle burn­ing red fuming nitric acid and aniline. Early plans called for use of centrifugal, turbine-driven pumps to feed the propellants. Since Aerojet had a turborocket under development, JPL thought it could draw on the nearby rocket firm’s experience to provide a pump for the Corporal. This design became the never-completed Corporal F. Corporal E used air pressurization, as had the WAC Corporal.10

Scaling the WAC Corporal engine up to a larger size proved chal­lenging. The first major design for a Corporal E engine involved a 650-pound, mild-steel version with helical cooling passages. Such a heavy propulsion device resulted from four unsuccessful attempts to scale up the WAC Corporal B engine to 200 pounds. None of them passed their proof testing. In the 650-pound engine, the cool­ing passages were machined to a heavy outer shell that formed a sort of hourglass shape around the throat of the nozzle. The injector consisted of 80 pairs of impinging jets that dispersed the oxidizer (fuming nitric acid) onto the fuel. The direction, velocity, and di­ameter of the streams were similar to those employed in the WAC Corporal A. The injector face was a showerhead type with orifices more or less uniformly distributed over it. It mixed the propellants in a ratio of 2.65 parts of oxidizer to 1 of fuel. The outer shell of the combustion chamber was attached to an inner shell by silver solder. When several of these heavyweight engines underwent proof test­ing, they cracked and nozzle throats eroded as the burning propel­lant exhausted out the rear of the engine. But three engines with the inner and outer shells welded together proved suitable for flight testing.11

On May 22, 1947, the first Corporal E with this heavyweight engine launched from the army’s White Sands Proving Ground. Its intended range was 60 miles, and it actually achieved a range of 62.5 (in one account, 64.25) miles. The second launch occurred on July 17, 1947, but the rocket failed to achieve enough thrust to rise significantly until 90 seconds of burning reduced the weight to the point that it flew a very short distance. On November 4, 1947, the third launch was more successful, but its propellants burned for only 43 (instead of 60) seconds before the engine quit. This reduced its range to just over 14 miles. Both it and the “rabbit killer" (the

second vehicle, so-called because it flew along the ground) expe­rienced burnthroughs in the throat area, the helical cooling coils proving inadequate for their purpose.12

Deciding that in addition to these flaws, the engine was too heavy, the Corporal team determined to design a much lighter-weight en­gine. Several engines combining features of the WAC Corporal B and 650-pound Corporal E combustion chambers all suffered burnouts of the throat area during static tests. Finally, a redesigned engine weighing about 125 pounds stemmed in part from an examination of the V-2, revealing that its cooling passages were axial (with no helix angle, i. e., they took the shortest distance around the combustion chamber’s circumference). Analysis showed the advantage of that ar­rangement, so JPL adopted it. The inner shell of the new engine was corrugated, and the outer shell, smooth. The shape of the combus­tion chamber changed from semispherical to essentially cylindrical, with the inside diameter reduced from 23 to 11 inches and the length shortened slightly, contributing to the much lighter weight.

It took two designs to achieve a satisfactory injector, the first having burned through on its initial static test. The second injector had 52 pairs of impinging jets angled about 2.5 degrees in the direc­tion of (but located well away from) the chamber wall. Initially, the Corporal team retained the mixture ratio of 2.65:1. But static tests of the axially cooled engine in November 1948 at the Ordnance – 150 California Institute of Technology (ORDCIT) Test Station in Muroc, Chapter 4 California (in the Mojave Desert above the San Gabriel Mountains and well north of JPL), showed that lower mixture ratios yielded higher characteristic velocities and specific impulses, as well as smoother operation. Thus, the mixture ratio was first reduced to 2.45 and then 2.2. Later still, the propellant was changed to stabi­lized fuming nitric acid (including a very small amount of hydrogen fluoride) as the oxidizer and aniline-furfuryl alcohol-hydrazine (in the percentages of 46.5, 46.5, and 7.0, respectively) as the fuel. With this propellant, the mixture ratio shifted further downward to 2.13 because of changes in the densities of the propellants. The resul­tant engine, made of mild steel, provided high reliability. Its suc­cess rested primarily upon its “unique configuration, wherein the cool, uncorrugated outer shell carrie[d] the chamber pressure loads, and the thin inner shell, corrugated to form forty-four axial cooling passages, [wa]s copper-brazed to the outer shell." Finally, the inside of the inner shell (the combustion chamber inside face) was plated with chrome to resist corrosion from the propellants.13

The sixth Corporal E launch took place on November 2, 1950. The missile experienced multiple failures. It landed 35.9 miles

downrange, about 35 miles short of projections. Later static tests revealed problems with a propellant regulator that had caused over­rich mixture ratios on both the fifth and sixth launches. Failure of a coupling had resulted in loss of air pressure. The radar beacon to provide overriding guidance in azimuth operated satisfactorily until failure of a flight-beacon transmitter some 36 seconds into the flight. The Doppler beacon never went into operation to cut off propellant flow at the proper moment because the missile failed to achieve the velocity prescribed, but also because the Doppler bea­con itself failed at 24 seconds after liftoff. As a final blow, all elec­tronic equipment failed, apparently from extreme vibration.14

On launch seven of the Corporal E in January 1951, the vehicle landed downrange at 63.85 miles, 5 miles short of the targeted im­pact point. This was the first flight to demonstrate propellant shut­off and also the first to use a new multicell air tank and a new air – disconnect coupling. These two design changes to fix some of the problems on launch six increased the reliability of the propulsion system significantly. However, although the Corporal performed even better on launch eight (March 22, 1951), hitting about 4 miles short of the target, on launch nine (July 12, 1951) the missile landed 20 miles beyond the target because of failure of the Doppler tran­sponder and the propellant cutoff system. The final “round" of Cor­poral E never flew. But the Corporal team had learned from the first nine rounds how little it understood about the flight environment of the vehicle, especially vibrations that occurred when it was op­erating. The team began to use vibration test tables to make the design better able to function and to test individual components before installation. This testing resulted in changes of suppliers and individual parts as well as to repairs before launch (or redesigns in the case of multiple failures of a given component.)15

The next 20 Corporals, with the airframes built by Douglas Air­craft (like the Corporal Es), received the designation Corporal I. Its first flight occurred on October 10, 1951. But the frequency regulator for the central power supply failed on takeoff, causing the missile to follow nearly a vertical trajectory. Range safety cut its flight short so that it would impact between White Sands and the city of Las Cruces, New Mexico. Flight 11 (referred to as round 12, counting the last Corporal E, which never launched) occurred on December 6, 1951. Before the launch, the army invited several companies to bid on production contracts as prime contractors. Ryan Aeronautical Company of San Diego manufactured the engines for both the air­frames built by Douglas and those from the new prime contractor, Firestone Tire and Rubber Company of Los Angeles. JPL received

the Firestone missiles and disassembled them for inspection. It then rebuilt them and performed preflight testing before sending them to White Sands for the actual flight tests. Then it sent comments to the manufacturer to help improve factory production. According to Clayton Koppes, however, the two major contractors and JPL failed to work together effectively. Meanwhile, between January and De­cember 1952, JPL launched 26 Corporals, including the first 10 of the Firestone lot as well as 16 produced by Douglas.16

Because of problems with the missile’s guidance system and engi­neering changes to correct them, a second production order to Fire­stone for Corporal missiles in late 1954 resulted in a redesignation of the missile as Corporal II. JPL retained technical control of the Corporal program throughout 1955, relinquishing it in 1956 while continuing to provide technical assistance to the army’s contrac­tors, including Firestone. Corporal II continued to have problems with its guidance/control system but also with propellant shutoff during firings of the missile by army field forces. Fact-finding in­vestigations and informal discussions on the parts of contractors, the field forces, Army Ordnance Corps, and JPL led to greater care by field forces personnel in following operational procedures. These eliminated shutoff problems when not violated. The army declared the Corporal to be operational in 1954, and in January 1955 the Corporal I deployed to Europe. Eight Corporal II battalions replaced 152 it during 1956 and the first half of 1957.17

Chapter 4 Although the Corporal was less powerful and had a shorter range than the V-2, the U. S. missile’s propulsion system had a higher spe­cific impulse (about 220 lbf-sec/lbm as compared with 210 for the V-2). In some respects, such as the axial nature of the cooling sys­tem and the use of Doppler radar for propellant cutoff, the Corporal had borrowed from the V-2. In most respects, however, the Ameri­can missile was an independent development, in some cases one that separately adopted features developed at Peenemunde after it was too late to incorporate them into the V-2. These included a showerhead injector and the use of hypergolic propellants. Both had been developed for the Wasserfall antiaircraft rocket, and a single injector plate later became a standard element in the construction of the rockets designed in Huntsville.18

Among the achievements of the Corporal was testing the effects of vibration on electronic equipment. The vibration tables used for this purpose may have been the first effective simulators of the flight environment in that area. Subsequently, both testing for the effects of vibrations and analysis of components and systems for reliability became standard practice in missile development.19

The engine itself was also a notable achievement. Although the idea for the axial direction of cooling flow came from the V-2, the overall engine was certainly original. It was both light and efficient, and even though there seems to be no evidence that its design influ­enced subsequent engines, it seems likely that propulsion engineers learned something of their art from it. Moreover, the early work of JPL in hypergolics transferred to Aerojet, later the contractor for the Titan II, which used storable liquid propellants that ignited on contact. This technology was also used in the Titan III and Titan IV liquid rockets, which employed direct descendants of the early hy – pergolic propellants Malina learned about in part from the navy in Annapolis. This was a significant contribution from both indige­nous U. S. research efforts during World War II. It illustrates one of the ways that technology transferred from one program to another in American rocketry. The borrowings from the V-2 exemplify a dif­ferent pattern of information flow.